DP2011-03

Sources of Global Heterogeneity in Retail Spending*

Nir KSHETRI

Ralf BEBENROTH

January 18, 2011

* The Discussion Papers are a series of research papers in their draft form, circulated to encourage discussion and comment. Citation and use of such a paper should take account of its provisional character. In some cases, a written consent of the author may be required.

Sources of Global Heterogeneity in Retail Spending

Nir Kshetri, The University of North Carolina at Greensboro, USA Ralf Bebenroth, Kobe University, Japan

 

Acknowledgement Part of this research was completed when Kshetri was visiting the Research Institute for Economics and Business Administration (RIEB) at Kobe University in October 2010 as a research fellow. He acknowledges the funding assistance of the RIEB.

1  

Sources of Global Heterogeneity in Retail Spending  

Abstract  

Economies worldwide vary greatly in terms of how much their consumers spend on various types of retail activities. The purpose of this paper is to examine how the regulatory characteristics as well as the natures and strategies of businesses are related to retail spending. We employed random effect time series cross sectional (TSCS) models linear in parameters for forty-eight economies using annual data for the 1999-2008 period. The results provided strong support that economic freedom, foreign direct investment (FDI) inflow and concentration of retail stores in an economy positively affect retail spending. We also found that tax and social security contributions as a proportion of the GDP is positively related to per capita grocery retail spending. A lack of data for a large number of economies, especially less developed ones potentially provides a limitation of this paper.

Keywords: Retail spending, FDI, economic freedom, time series cross sectional (TSCS) models

 

1. Introduction

Economies worldwide vary greatly in terms of how much their consumers spend on various

types of retail activities. According to Euromonitor International, for instance, per capita

retail spending for 2006 varied from US$110 for Pakistan to US$9,338 for Norway. Likewise,

Poland has 133 square meters of shopping-center space per 1,000 inhabitants, compared with

Norway’s 632.5 square meters (Kilbinger, 2007).

Consumers may save part of their disposable income and invest in various assets. The

remaining part of the income is spent on retail goods as well as on the demand of non-retail

goods such as entertainment, educational opportunities, health services, and housing. Retail

spending can be further divided into grocery retails (e.g., food products) and non-grocery

retails, which can be further broken into hard goods (e.g., appliances, electronics, furniture,

sporting goods, etc.) and soft goods (e.g., clothing, apparel, and other fabrics). One important

question then is how economies worldwide vary in terms of the proportion of the income

spent in various categories of retailing.

2  

With the rapid rate of retail internationalization, the study of retailing is rapidly taking

off in international business and marketing literatures (Alexander and Myers, 2000). It is,

however, difficult to draw any wide-ranging conclusions from these studies concerning

international variation in retail spending. In this paper, we would thus argue that the drivers

of retail spending, especially in the global context, while well documented, are only partially

understood.

In prior theoretical and empirical research, scholars have noted that

internationalization of the retailing industry is a “complex and relatively poorly understood

phenomenon" (Vida and Fairhurst, 1998, p. 141). There is growing recognition that the

internationalization of retailing is a phenomenon that is highly understudied in the academic

literature (Coe, 2004; Wrigley, 2000a). Prior researchers have noted the importance of more

research in this area (Alexander and Myers, 2000), especially on the quantification front.

Vida et al. (2000) noted that empirical research in international retailing “remains scarce and

lacks integration”.

Prior researchers have also suggested that consumers behave differently for grocery

(e.g., foods) and non-grocery (e.g., clothes, books, CD) retailing (Næss, 2006). What is not

clear is what factors explain such differences.

Econometric models in particular will provide valuable insights into the economic

geography of retailing (Anderson, Chatterjee and Lakshmanan, 2003). Such models allow for

policy simulations with respect to changes in marketing actions. Such a line of thought is

also echoed in more recent perspectives on marketing research (Bronnenberg, Rossi and

Vilcassim, 2005; Chu, Chintagunta and Vilcassim, 2007; Franses, 2005; Kshetri, Williamson

and Schiopu, 2007). Researchers have also emphasized the importance of going beyond

multinationals’ point of entry into a foreign market and studying the temporal and spatial

dimensions of retail internationalization (e.g., Clarke and Rimmer, 1997; Wrigley, 2000b).

3  

Most obviously, at the country level, per capita income is an important determinant of

retail spending. Yet, having said this, it is apparent, too, that factors other than income

explain to a significant proportion of the variation in retail spending. To substantiate this

claim, we now call attention to Table 1. As we can see in the table, Azerbaijan’s per capita

retail expenditure is greater than Belarus’ notwithstanding a much lower income level of the

former compared to that of the latter. Similar relationships can be observed if we compare

corresponding figures for Brazil and Bulgaria.

To more fully understand international variation in the development of the retail

industry, in this paper, we examine how the policy environment as well as the nature and

strategy of retailers drive retail spending in the economy. In the remainder of the paper, we

first provide a literature review and develop some hypotheses. Following is a section on

methods. Then, we discuss the results. The final section provides conclusion and implications.

Insert Table 1 about here

2.0 Literature Review and Hypotheses Development

Foreign direct investment (FDI)

A study conducted in Mexico showed that MNCs' presence accelerates retail modernization

(Durand, 2007). Similarly, McKinsey Global Institute’s study of retailing in China, India,

Brazil, and Mexico found that foreign companies improve efficiency and productivity. They

do so by bringing capital, technology, and management skills and forcing inefficient domestic

companies to either improve their operations or to exit (Farrell, 2004).

In a discussion of the FDI effect on the development of the retail industry, the most

relevant issues concern multinational retailers’ innovative practices (Chan and McNeal, 2006;

Davis and Sensenbrenner, 2000), which are likely to stimulate retail industry in the host

country. Especially in developing countries, distribution and retailing channels are

4  

fragmented (The Economist, 1991). These economies are not prepared to transform these

services on their own (Keren and Ofer, 2000, 2002). Compared to local firms, global retailers

arguably have better ways of managing supply chains and marketing activities. Global

retailers possess assets based on conceptual and technological advantages (Alexander and

Myers, 2000). For instance, Samiee (1995) lists three areas in which U.S.-based retailers

have considerable advantages: 1) experience in the use of information technology (IT); 2)

strong customer service; and 3) possession of advanced and sophisticated marketing skills in

various areas1.

International retailing firms possess systems and processes that are largely lacking in

developing countries (Samiee, Yip and Luk, 2004). Multinationals (MNCs) in supermarkets

and other sectors engage in 'proactive fast-tracking strategies' to change the 'enabling

conditions' of entry and growth via mechanisms such as procurement system modernization

and local supply chain development (Reardon, Henson and Berdegué, 2007). For instance,

FDI of Skoda and Volkswagen in the Czech Republic stimulated new supply networks

through backward linkages (Bohata 2000; Keren and Ofer, 2000, 2002; The Economist,

2001). MNCs in the retail sector also engage in 'follow sourcing' strategy, that is, they

encourage the entry of other MNCs in logistics and wholesale firms, which are parts of the

former’s supplier network in its home country (Reardon, Henson and Berdegué, 2007).

Volkswagen’s operation in the Czech Republic discussed above is an example of follow-

sourcing. Multinational retailers also engage in multi-network-sourcing, that is, they use

national, regional and global networks to source (Coe and Lee, 2006; Reardon et al., 2007).

Compared to local firms, global retailers can also manage their marketing mix related

variables in a better way. For instance, an important role of advertising is to “manipulate the

desires and habits of the common man” (Gagnon, 2007) and few local firms in a primitive

5  

market can do so more skillfully than global MNCs on this front. Multinational retailers are

also likely to have superior goods to sell (Haslach, 1992; Lipman, 2007).

An especially interesting case of a rapid development of the retail industry is that of

the post-WTO China. By 2013, retail sales in China, excluding autos, are estimated to be $1.6

trillion. In A.T. Kearney's ninth annual study of global markets with potential for retail

development, China ranked first (WWD 2010). After the WTO accession in 2001, China

allowed foreign retail participation in joint ventures (JVs) in the five special economic zones

(Hainan, Shantou, Shenzhen, Xiamen and Zhuhai) and in the cities of Beijing, Dalian,

Guangzhou, Qingdao, Shanghai and Tianjin (Country Commerce 2006). In April 2004,

China announced that it would allow 100%-foreign-owned companies to engage in retailing

and wholesaling. From December 2004, the Chinese Ministry of Commerce (MOFCOM)

started approving applications from companies (Country Commerce 2006; Chandler, 2005).

Over 35 of the global top 50 retailers entered into the country by 2007 (King, 2007). China

has been the world’s biggest market for cars since 2009. In this regard, it is worth

emphasizing that car prices in China dropped by over 30 percent during 1995-2001, while

overall consumer prices grew by 10 percent during the same time period (Farrell, 2004).

Global retailers’ footprints across the world economy are getting bigger. To take one

example, Wal-Mart's international division is growing faster than the company's U.S.

business (Los Angeles Times, 2007). Likewise, 60% of the U.K. based food retailer Tesco’s

outlets are outside its home market (Kilbinger, 2007). Commenting on the nature of “a virtual

headlong rush” of many retailers, a Chain Store Age (1997) article put the issue this way:

“The change isn't simply about numbers. For one thing, those first international efforts were

often limited to given regions--U.S. retailers entering Canada, pan-European retailing, or

Asian stores branching out within the Pacific Rim. Today, however, retailers are increasingly

crossing oceans and thinking, sourcing, and selling globally”.

6  

An increasing proportion of foreign investment in developing economies is going to

distribution and retail sectors (Samiee, Yip and Luk, 2004). One estimate suggested that the

top 25 global retailers would account for 40 percent of the world's retail spending by 2009

(Christian Science Monitor, 1999). Likewise, seven multinationals dominated the global

retail market for hot drinks, which was estimated at US$75 billion in 2004 (Business Wire

2005).

Factors related to the retail sector as well as the general business environment (e.g.,

saturation of the home markets, higher growth opportunities in foreign countries, increasing

homogeneity of consumer markets worldwide, and the opening of emerging markets that had

underdeveloped retail sector) have led to a rapid internationalization of retailers (e.g.,

Alexander 1997; Kumar 1997; Sternquist 1997; Vida et al. 2000).

The retail industry and spending in this industry in many developing countries have

received a big boost due to global retailers’ innovative practices (Chan and McNeal, 2006;

Davis and Sensenbrenner, 2000). For instance, FDI and joint ventures have replaced the old

distribution and production network systems of central planning in Central and Eastern

European (CEE) economies (Keren and Ofer, 2002).

Similarly, China’s advantage over India in the development of the retail sector is

partly due to foreign retailers’ higher investments in the former. China also lowered its tariffs

faster than India (Chu and Magnier, 2006). Chinese retail spending is rising meteorically. In

India, on the other hand, there are over 12 million small retailers in the unorganized sector

(Dutt, 2005), which accounted for 97 percent of retail revenue in 2005 (Rai, 2006). More

recent estimates suggest that the organized sector accounts for 8 % of urban Indian retail

spending and even lower proportion in rural India (Kazmin, 2010).

India allows only single-brand foreign retails to enter into the country (Kazmin, 2010).

To conform to restrictions protecting Indian retailers, Wal-Mart was required to form an

7  

alliance with Bharti's retailing division (Chandler, 2007). Indian politicians have argued that

global retailers may drive local players out of business (Rai, 2006). Recent legislation relaxed

some of these rules.

Foreign retailers’ effects are even more readily apparent in post socialist economies,

in which decades of socialism suppressed the market. In particular, retailing, and marketing

more broadly, were not part of the DNA of these economies. For instance, it is argued that

Russian companies are run by "antiquated modes of organization", which are characterized

by inadequate marketing and sales skills, lack of quality control, and lack of incentives and

teamwork (The Economist, 1999). In these economies, consumers lack choices. Starting the

late 1980s, retail multinationals (e.g., McDonald and IKEA) 2 started entering these

economies through FDI. Their entry arguably contributed to the development of retail

services in these economies (Keren and Ofer, 2000).

Prior researchers have noted that domestic services developed during the early phase

of the transition lacked customer orientation and consisted of inferior know-how, technology,

management and organization and provided low quality services (Keren and Ofer, 2000,

2002). Overall, foreign retailers’ lower prices and better selection often tend to stimulate

consumer demand and industry growth (Farrell, 2004). Beyond all that, there are also indirect

effects related to the growth in national income improved productivity and output (Farrell,

2004). The above leads to the following:

H1a: The level of per capita foreign direct investment is positively related to per capita retail spending. H1b: The level of per capita foreign direct investment is positively related to per capita grocery retail spending. H1c: The level of per capita foreign direct investment is positively related to per capita non-grocery retail spending.

Concentration of retail stores

8  

The idea of retail convenience (Seiders, Berry and Gresham 2000) deserves mention. Retail

convenience is related to a consumer’s time and effort costs in shopping (Beauchamp and

Ponder, 2010). For one thing, retail convenience is related to consumer’s time savings (Yale

and Venkatesh 1986). Seiders et al. (2000) propose four dimensions of convenience relevant

to retailing: access, search, possession, and transaction.

We would argue that access and transaction dimensions and consumer’s time savings

(Seiders et al. 2000; Yale and Venkatesh 1986) are affected by the concentration of retail

stores. There are reasons to expect that concentration of retail stores stimulate retail spending

by facilitating convenience. As retailers face more competitors, they are forced to decrease

the prices, which would help them attract price-conscious consumers. There is some evidence

to suggest that an increase in the supply of stores has led to downward pressures major retail

firms’ prices (Harris and Vega, 1996). In China and Russia, foreign as well as and local

retailers have expanded into the urban as well semi-urban areas, which have stimulated the

development of the retail industry (Khanna et al. 2005). Based on above discussion, the

following hypotheses are presented:

H2a: Concentration of retail stores is positively related to per capita retail spending.

H2b: Concentration of grocery retail stores is positively related to per capita grocery retail spending.

H2c: Concentration of non-grocery retail stores is positively related to per capita non-grocery retail spending.

Economic freedom

Of equal importance in the discussion of the development of the retail industry that follows

below is the degree to which the government facilitates an “unfettered”’ operation of retailers.

There are a number of pieces of evidence indicating that economic freedom or openness

drives the development of retailing (Dickson, 2000). For instance, in recent years, global

9  

consumer goods retailers have been encouraged by news that the Chinese government's

interference in the foreign-exchange market may gradually decline (WWD 2010).

The regulatory environment or a lack of economic freedom has created various

obstacles to the development of the retailing industry in some countries. One of the major

obstacles centers around regulations related to store operating hours (Samiee 1995). In

Europe, while Sweden and the U.K. have significantly liberalized trading hours, countries

such as Austria, Denmark, Finland, Germany and Norway still impose a number of

restrictions on shopping hours. There are virtually no major retail activities in these countries

on Sundays. Not long ago in Germany, stores were required to close by 8 PM on weekdays,

and by 2 PM on Saturdays (Shy and Stenbacka, 2008).

To put in the historical context, Landes (1969, p. 52) comments how the openness

drove the development of the British retailing industry in the 19th century: "British

shopkeepers were relatively free of customary or legal restrictions on the objects or character

of their activity. They could sell what and where they would; and could and did compete

freely on the basis of price, advertising and credit. If most shopkeepers continued to haggle,

many followed the lead of the Quakers in selling at fixed, market prices. In so far as such

methods prevailed, they conduced to a more efficient allocation of economic resources and

lower costs of distribution".

In many emerging economies, government controls on ingredients of the retailing

system have hindered the growth of this sector. Foreign as well as domestic firms may face

various constraints. Walker and Etzel’s (1973) survey found that internationalizing

franchisors faced legal and other problems. In April 2007, the Russian government outlawed

foreign migrant workers in the Russian retail markets. Another restriction requires that at

least half of all salespeople sell their own produce (Lipman, 2007). Ex-President Putin argued

that in Russia, foreigners have “long ruled the roost” (Nikolayeva, 2006). Likewise, an Indian

10  

law enacted in 1966 banned farmers from dealing directly with retailers and forced them to

sell through licensed middlemen, called “mandis” (Robinson, 2007).

Domestic retailers have played an important role in the development of retail

industries in economies such as Japan, Mongolia, Romania and Russia (Gumbel, 2006; Reid,

1995). A common thread that runs through these economies is that domestic as well as

foreign firms’ involvements in retailing are facilitated by reform measures. In the 1990s,

Japan proposed deregulation plans for retailing and distribution (Mallaby et al. 1998). The

Large-scale Retail Store Law was amended in 2006. However, companies with the intention

of entering the Japanese retail market still complained about disadvantages against existing

companies (Buckley, 2007, 20). Likewise, many developing economies have launched

economic reform programs to free up markets due to pressures from the World Bank (The

Economist, 1995).

In the Soviet Union, the retail industry was state-owned (Hilton, 2004). Note too that

service activities such as retailing, wholesaling and consulting go against the Marxist

ideology (Goldman, 2006). Following the late 1980s, in Russia, big supermarket operators

such as Seventh Continent and Pyaterochka were publicly listed on the stock exchange

(Gumbel, 2006). The Russian retailing industry is also consolidating. In 2006, Pyaterochka

merged with rival Perekriostok to form Russia's biggest food retailer, with almost 900 stores

and sales of $2.4 billion (Gumbel, 2006). Note that Russia’s retail trajectory is significantly

different from those of other East European countries such as Poland and the Czech Republic.

In Poland and Czech Republic, the retail industry is driven by Western retailers, which are

hesitant to enter Russia (Gumbel, 2006). Likewise, factors such as easier borrowing and

lower taxes—some indicators of economic freedom- are driving retail spending in Romania

(BMI, 2006).

11  

In the late 1980s, private retail businesses in Poland were encouraged to compete with

the state (Diehl, 1988). This difference is powerfully illustrated in the differential

development of supermarket outlets in Poland and Russia. In the late 1990s, supermarket

outlets had a 1% market share in Russia compared with 18% in Poland (The Economist,

1999).

Another way in which the government can restrict the development of the retail sector

is through control in locations of retail activities. Unlike in many Western cities (Wang and

Zhou 1999), the location of central business districts cannot be located in city centers in

China (Yan, 1995, p. 481). Retailing and commercial activities are relatively less emphasized

in Chinese city centers (Wu, 1993). For instance, since 1949, the Chinese government has

emphasized Beijing’s role as the national political and cultural centre and restricted

commercial and industrial activities through city plans (Hu, 1993). Such a “politically

motivated stigmatization” of city centers seems to be apparent in other cities in former

socialist countries (Bater, 1980, pp. 125-130; French and Hamilton, 1979, p. 318; Altrock,

1998; Borchert, 1979, 1995; Kulke, 1992, 1996, 1997). Thus:

H3a: The level of economic freedom is positively related to per capita retail spending.

H3b: The level of economic freedom is positively related to per capita grocery retail spending.

H3c: The level of economic freedom is positively related to per capita non-grocery retail spending.

Tax and social security contributions

It is reasonable to assume that retail spending decreases when tax and social security

contributions increase. This is because, a high level of tax and social security contributions

leads to a decline in consumers’ spending power.

That said, in some economies, tax and social security contributions have been a major

driver of consumers’ grocery retail spending. In 2004, the U.S. spent $24.6 billion on the

Food Stamp Program (FSP) and $34.7 billion on Supplemental Security Income (SSI)

12  

benefits (Spar 2006, cited in Trenkamp and Wiseman, 2007). In 2007, FSP served 26 million

Americans (Congressional Food Stamp Challenge, 2007). This means that an average FSP

beneficiary family received over $900 annually. This is significant if we consider the fact that

average yearly expenditures on food in U.S. urban household was $2,207 in 2004 (Blisard

and Stewart 2007). There are additional mechanisms such as the Earned Income Tax Credit

(EITC) that are available to help the low-income families. The EITC is arguably one of the

main Federal programs to subsidize the working low-income individuals.

A final point that should be taken into account concerns non-grocery retail’s

dominance in the retail industry. For instance, grocery retail spending is estimated to account

for only 42% of total retail spending worldwide and only 34% in Gulf Co-Operation Council

markets (Branston, 2009). We thus expect that the relation between tax and social security

contributions and retail spending is more likely to be described by non-grocery retail

spending than by grocery retail spending. Thus:

H4a: The level of tax and social security contributions is negatively related to per capita retail spending. H4b: The level of tax and social security contributions is positively related to per capita grocery retail spending. H4c: The level of tax and social security contributions is negatively related to per capita non-grocery retail spending. Concluding hypothesized relationships

From the above discussions it should be noted that the major hypotheses concerning the

drivers of per capita retail spending (grocery, non-grocery as well as total) are that: (1) it is

positively related to per capita FDI inflow; (2) it is positively related to economic freedom;

(3) it is positively related to the concentration of retail stores. In addition, we hypothesize that

(4) per capita grocery retail spending is positively related to tax and social security

contributions as a proportion of GDP.

13  

3. Methods

Our unit of analysis is an economy.

Dependent variables

The dependent variables used in the analyses are per capita retail spending excluding sales

tax (PCRetail), per capita grocery retail spending excluding sales tax (PCGRetail) and per

capita non-grocery retail spending excluding sales tax (PCNGRetail).

Explanatory variables

Explanatory variables used in this study include per capita GDP measured at purchasing

power parity (US$) (PCGDPPPP), retail sites/outlets per 1000 people (PCRetailsites),

grocery retail sites/outlets per 1000 people (PCGRetailsites), non-grocery retail sites/outlets

per 1000 people (PCNGRetailsites), foreign direct investment inward stocks per person

(PCFDIIS)3, economic freedom (score) (ECFR), and Tax and social security contributions

as % of gross income (TSSC).

Control variables

As control variables, we used per capita GDP measured at purchasing power parity (US$)

(PCGDPPPP), the population density (people per sq km) (PD), savings ratio as a percentage

of disposable income (SR), and businesses’ advertising spending divided by the population

(PCTAFA).

Data sources

Prior to analyzing the data, we developed hypotheses regarding potential sources of

heterogeneity in the development of the retail industry. We analyzed 48 economies for which

data on dependent and independent variables were available. Table 2 and Table 3 present

descriptive statistics and correlation matrix respectively for variables for 2006.

Insert Tables 2 and 3 about here

14  

Most of the data related to retail spending and related indicators were obtained from

Euromonitor publications. We used the Economic Freedom Index calculated by The

Heritage Foundation, which formulates this index for 161 economies. Before the 2007

estimates, the index was based on 9 freedom variables related to regulation, trade, fiscal,

government, monetary, investment, financial, property rights, and corruption. Starting 2007,

labor freedom was added to the list4. These variables capture difficulties of economic and

political natures that are likely to face retailers in an economy (Colla and Dupuis, 2002).

It is worth noting that there are five major constraints related to the use of any

international secondary data: accuracy, age, reliability, lumping and comparability (Kotabe

and Helsen, 2001). Kotabe (2002) argues that Euromonitor, despite its reliance on various

sources, addresses the first four constraints. Regarding comparability, it is also important to

note that this constraint is mainly a consequence of a lack a common and shared

understanding of a concept (e.g., social capital) across countries (Harper, 2002). This

problem is compounded by different languages used in the surveys for measuring the

concepts. Since the data used in this paper represent actions rather than attitude, felling or

intention and have straightforward operationalization, international comparability doesn’t

seem to be a problem. Kotabe (2002) observes: “Usually, the measurement quality of data

collected from reputable data sources such WMDS [Euromonitor’s World marketing data and

statistics] do not get challenged in the blind review process” (p. 174). Euromonitor data have

been used in several studies including Coulter et al. (2003), Dixon and Karboulonis (1999),

Ein-Dor et al. (1997), Ganesh (1998), Ganesh et al. (1997). Likewise, the Heritage

Foundation data have been used in studies such as Rose (2004), Vega-Gordillo and Alvarez-

Arce (2003), Hanson (2003), Lau and Lam (2002).

15  

Statistical Analysis

We employed time series cross sectional (TSCS) models to analyze the data. We analyzed

annual data for 10 years (1999-2008). Note that TSCS models are designed to overcome the

limitations of usual linear models. When pooling data, it is highly likely that one or more

assumptions of the usual linear model may be violated. Fomby et al. (1984, p. 337) point out

several such possibilities. First, the error terms in a pooled model may be “heteroskedastic,

autocorrelated and may exhibit contemporaneous correlation” which make generalized least

square technique inappropriate. Second, the parameters of the data-generating process may

differ from observation to observation. The reactions of different individuals may be different

to changes in explanatory variables and the reactions may also change over time. TSCS

models allow for differences in behavior over cross sectional units and also the differences in

behavior over time for a given cross section. In this way, such models are likely to be

consistent with the way the data were generated (Fomby et al., 1984). Problems related to

such models include the selection of the most efficient estimation procedures and testing of

hypotheses about the parameters.

We employed the TSCS models in the following form:

),1(2

PCRetail 1 axK

kitkitkititit

),1(2

PCGRetail 1 bxK

kitkitkititit

),1(2

PCNGRetail 1 cxK

kitkitkititit

Where,

16  

PCRetail: per capita retail spending excluding sales tax.

PCGRetail: per capita grocery retail spending excluding sales tax .

PCNGRetail: per capita non-grocery retail spending excluding sales tax.

1it is the dummy variable for the ith country for the tth time period and kit (k 2) are the

slopes. Xkit (k 2) is the value of the predictor Xk for the ith country in time t.

Several factors need to be taken into consideration in selecting the best TSCS model.

The first is the choice between fixed and random effect models. For the fixed effect (or

dummy variable) model, the intercept term 1it in (1) can be written as

1it = i + t (2),

where i are the country “dummies” and t are the time “dummies”. The dummy variable

model, however, eliminates a major portion of the variation among explained as well as

explanatory variables if the between-country and between-time period variation is large

(Maddala, 1971). Additional problems include a loss in a substantial number of degrees of

freedom and a lack of meaningful interpretation of the dummy variables (Maddala, 1971).

These problems can be overcome by treating i and t as random in which case only

two parameters, the mean and the variance of the 's (and similarly for 's), are estimated

instead of N+T parameters in dummy variable models, where N is the number of cross-

sections and T is the number of time periods. The procedure of treating i and t as random

can be rationalized by arguing that the dummy variables do in effect represent some

ignorance – just like it. Maddala (1971) argues that this type of ignorance, or “specific

ignorance,” can be treated in the same manner as it. Therefore, the residual can be written as:

uit = i + t+ it (3).

In TSCS models, two considerations, logical and statistical, may determine the choice

of specification—fixed vs. random (Hausman 1978). The logical consideration is whether 1it

17  

can be considered random and drawn from an independently and identically distributed (IID)

distribution (Hausman, 1978, p. 1263). The statistical consideration is whether 1it’s satisfy

“di Finnetti’s exchangeability criterion” (p. 1263), a necessary and sufficient condition for

random sampling. If these conditions are satisfied, then random model can be more

appropriate than a fixed model. To empirically test the statistical consideration, we estimated

the fixed effect model5 for 48 cross-sections for which “complete” data for the period under

consideration were available. Then we calculated the correlation between the country specific

fixed effects and time specific fixed effects with other country specific factors or regressors

(Tables 4 and 5). As tables 4 and 5 indicate, of the 42 Pearsonian coefficients, only one is

significant. Since most of the Pearsonian correlation coefficients were insignificant, it

became clear that random effect TSCS models are more appropriate for the given data set

than fixed effect TSCS models.

Insert Tables 4 and 5 about here

After knowing the appropriateness of the random effect TSCS models over fixed

effect ones, the next step would be to select the most appropriate random effect model. In the

pooled data on Internet diffusion, it is reasonable to expect heteroskedasticity [i.e. E(uit2) =

ii], contemporaneous correlation or spatial heterogeneity [i.e. E(uitujt) = ij] (Anselin 1987),

and autoregression [i.e. uit = iui,t-1+eit]. Among the three most commonly used estimation

procedures for random effect TSCS models–Fuller-Battese, Da Silva and Parks– the Fuller-

Battese (Fuller and Battese 1974) takes only heteroskedasticity into account while Da Silva

(1975) considers heteroskedasticity and autoregression. Parks (1967) method, on the other

hand, takes heteroskedasticity, autoregression as well as contemporaneous correlation into

account and hence appears to be the most appropriate method to study the multi-country

retailing. We estimated the parameters by using Parks method (SAS Institute Inc., 1999).

18  

4.0. Results and Discussion

TSCS results are displayed in Tables 6 and 7. For PCFDIIS data were not available for 2007

and 2008. We thus estimated the TSCS models for 8 years (1999-2006) for models that

included PCFDIIS as an explanatory variable (Table 6).

It is important to note that conventional measures of R2 are inappropriate for TSCS

models (SAS Institute, 1999: 1136). We thus did not report R2 values for the models.

Hypothesis 1a, 1b and 1c predicted that the level of per capita foreign direct

investment has a positive effect on retail spending as well as on grocery retail spending and

non-grocery retail spending. The TSCS results (Tables 6) provide support for H1b. We,

however, found no supports for H1a and H1c. This may be due to the fact the FDI is more

prevalent in the grocery sector than in the non-grocery sector. For instance, during the early

1990s, in Czech Republic, while some foreign mass merchandisers entered the country,

foreign investment was more readily apparent in food stores (Pellet, 1995).

The TSCS results (Tables 7) indicate that H2a, H2b and H2c are supported (t = 2.78, p <

0.01 for overall retail spending, and t = 6.68, p < 0.001 for grocery retail spending, and t =

3.14, p < 0.01 for non- grocery retail spending). It is also important to note that in the

regression model with PCNGRetail as the dependent variable (Table 6), PCFDIIS acts as a

suppressor variable, which is defined as a variable that increases the predictive validity of

another variable (in this case, PCNGRetailsites) by its inclusion in the model (Conger, 1974).

Results similarly provide strong support for H3a, H3b, and H3c. As above, in the

regression model with PCNGRetai as the dependent variable (Table 6), PCFDIIS acts as a

suppressor variable by increasing the predictive validity of ECFR by its inclusion in the

model (Conger, 1974). Finally, as predicted by the hypothesis 4b, the variable related to tax

and social security contributions as a proportion of gross income (TSSC) has a significant

positive effect on grocery retail spending (t= 2.27, p < 0.05 in Table 6 and 7.02, p < 0.001 in

19  

Table 7). This variable, on the other hand, has no significant effect on non-grocery retail

spending (Tables 6 and Table 7). Thus, hypothesis 4c is not supported. However, contrary to

hypothesis 4a, TSSC has a positive effect on overall retail spending.

Overall, most of the hypothesized effects are statistically significant and in the

expected directions. The results indicate that the drivers of the retail industry for necessity

items (e.g., grocery) are slightly different from those for luxury items or non-necessity items.

Insert Tables 6 and 7 about here

5.0. Conclusion and Implications

In this paper, we examined the sources of heterogeneity in retail spending in economies

across the world. To the authors’ knowledge, this is the most comprehensive cross- national

comparative study on the drivers of the retail industry. A major contribution of this research

is to confirm by clear and convincing evidence that economies that exhibit a higher degree of

openness to outside investment (a higher FDI inflow) and with having a higher economic

freedom are likely to experience a higher rate of development of their retail industry. Our

study also indicated that concentration of retail stores drives retail spending and even an

increase of tax and social security contributions lead to a higher grocery retail spending.

In spite of the above contributions, some limitations accompany these analyses. Prior

researchers have recognized several statistical issues and challenges associated with

international secondary data. For instance, in some cases, retail spending is underestimated

because the official figures may omit market stalls (Consumer Eastern Europe, 1992).

Another limitation of this study concerns the focus mainly on regulation and business related

variables and the supply side of the equation. A further limitation concerns the omission of a

large number of economies, mainly the low-income ones due to data unavailability.

20  

Further research is needed to validate, extend, refine, and assess the generalizability

of the comprehensive model of retail spending presented in this paper. Additional variables

such as availability of services and supports to consumer can be added.

Another intriguing avenue for future research is to examine if the saving rate has

differential impacts on various categories of retailing spending. In this paper, we used saving

rate as a control variable. For instance, consumers may save part of their disposable income

in order to purchase consumer durable items such as a car at some later date. There are thus

reasons to believe that saving rate is likely to stimulate spending in hard goods (e.g.,

appliances, electronics, furniture, sporting goods, etc.) compared to food products or soft

goods (e.g., clothing, apparel, and other fabrics). It is worth examining this issue in greater

detail.

Finally, whereas the quantitative models employed in this paper placed emphasis on

insuring "tightness of control", future research can pursue "richness in reality" (Mason, 1988)

of the global retailing industry by using qualitative methods such as interpretive consumer

research. Qualitative methods provide the basis for elaborations, interpretations and new

ideas (Morse, 1992) and will further deepen our understanding of how consumers can make

use of retailing.

21  

Table 1: A comparison of retail spending and related indicators in selected emerging economies

 

Per capita retail expenditures (US$, 2006)

GDP per capita (US$, 2006)

A measure of Economic Freedom as derived by The Heritage Foundation (2006)

Per Capita Foreign Direct Investment (US$, 2005)

Population (million, 2006)

Azerbaijan 432 1,426 53.96 198 8,569 Belarus 414 3,776 48.54 31 9,784 Brazil 697 5,640 61.71 81 189,323 Bulgaria 1,058 4,064 64.29 288 7,676 Source: Calculated from Euromonitor International

22  

 

Table 2: Descriptive Statistics for variables (1999-2006, No. of observations: 384)

Mean Std Dev Minimum Maximum

PCRetail (US$) 3065.34 2491.58 102.5 9742.65

PCGDPPPP(US$) 18748.74 11520.99 1307.56 50034.32

TSSC (%) 19.86 9.9 2.9 42.8

ECFR (score) 65.67 9.74 42.68 89.97PD (people per sq km) 405.12 1279.48 2.5 6926.4

PCTAFA (US$) 144.37 135.14 1.39 775.48

SR (%) 9.08 8.48 -15.6 33.6PCRetailsites (per 1000 people) 7.63 3.32 1.56 17.19

PCGRetail (US$) 1388.79 1151.17 73.68 4860.63PCGRetailsites (per 1000 people) 3.73 2.41 0.72 15.29

PCNGRetail (US$) 1676.55 1396.57 28.83 5037.74PCNGRetailsites (per 1000 people) 3.89 2.35 0.45 10.96

Table 3: Correlation matrix for variables in 2006 (N = 48)

PCRet

ail PCGDPPPP

TSSC

ECFR

PD PCTAFA

SR PCRetail

sites PCRe

tail PCGRetailsites

PCNGRetail

PCTAFA 0.87 ***

0.88 ***

0.52 ***

0.70 ***

0.19

SR -0.05 0.15 -0.26 0.04 +

0.51 ***

-0.01

PCRetailsites -0.20 -0.21 -0.29 *

-0.18 0.006 -0.18 -0.21

PCGRetail 0.98 ***

0.81 ***

0.65 ***

0.56 ***

-0.13 0.82 ***

-0.10 -0.18

PCGRetailsites

-0.48 ***

-0.54 ***

-0.51 ***

-0.38 **

-0.06 -0.44 **

-0.10 0.74 ***-0.44

**

PCNGRetail 0.98 ***

0.91 ***

0.55 ***

0.69 ***

0.07 0.87 ***

0.006 -0.21 0.92 ***

-0.50 ***

PCNGRetailsites 0.28 + 0.32 * 0.17 0.18 0.08 0.26 + -0.19 0.61 *** 0.25 + -0.07 0.29 *

+Significant at 0.1 level, *Significant at 0.05 level, ** Significant at 0.01 level, ***Significant at 0.001 level

23  

Table 4: Pearsonian correlations between country and time specific fixed effects with regressors

Variable Pearsonian correlation coefficient with country specific fixed effect (p-value)

Country effects for PCRetail as dependent variable

Pearsonian correlation coefficient with country specific fixed effect (p-value)

Country effects for PCGRetail as dependent variable

Pearsonian correlation coefficient with country specific fixed effect (p-value)

Country effects for PCNGRetail as dependent variable

ECFR -0.029 (0.843) -0.038 (0.798) -0.024 (0.870)

TSSC 0.203 (0.166) 0.214 (0.144) 0.188 (0.201)

PCFDIIS -0.088 (0.553) -0.089 (0.550) -0.091 (0.539)

PCRetailsites -0.057 (0.698)

PCGRetailsites -0.096 (0.515)

PCNGRetailsites -0.014 (0.925)

PCTAFA -0.009 (0.949) 0.003 (0.983) -0.027 (0.856)

SR -0.058 (0.694) -0.058 (0.696) -0.062 (0.675)

PCGDPPPP -0.001 (0.995) 0.015 (0.918) -0.022 (0.882)

PD -0.079 (0.592) -0.089 (0.549) -0.073 (0.623)

+Significant at 0.1 level, *Significant at 0.05 level, ** Significant at 0.01 level, ***Significant at 0.001 level

24  

Table 5: Pearsonian correlations between time specific fixed effects with regressors

Variable Pearsonian correlation coefficient with time specific fixed effect (p-value)

PCRetail as Dependent variable

Pearsonian correlation coefficient with time specific fixed effect (p-value)

PCGRetail as Dependent variable

Pearsonian correlation coefficient with time specific fixed effect (p-value)

PCNGRetail as Dependent variable

ECFR -0.191 (0.597) -0.192 (0.595) -0.244 (0.497)

TSSC -0.459 (0.182) -0.507 (0.135) -0.401 (0.251)

PCRetailsites 0.483 (0.158)

PCGRetailsites 0.787 (0.007)**

PCNGRetailsites -0.052 (0.887)

PCTAFA 0.316 (0.374) 0.323 (0.362) 0.256 (0.475)

SR 0.146 (0.688) 0.155 (0.668) 0.149 (0.680)

PCGDPPPP 0.135 (0.709) 0.139 (0.701) 0.085 (0.815)

PD 0.028 (0.938) 0.026 (0.942) -0.017 (0.963)

+Significant at 0.1 level, *Significant at 0.05 level, ** Significant at 0.01 level, ***Significant at 0.001 level

25  

Table 6: TSCS analysis (N = 48)

Model 1 (Dependent variable: PCRetail)

Model 2 (Dependent variable: PCGRetail)

Model 3 (Dependent variable: PCNGRetail)

Intercept -970.825 (3.23) ** -298.742 (4.99) *** -311.635 (3.07) ** ECFR 5.124332 (1.91) + 5.773793 (12.39) *** 1.742168 (1.28) TSSC 16.46031 (4.54) *** 2.944875 (7.02) *** 2.328461 (1.22)

PCRetailsites 48.49386 (2.78) ** PCGRetailsites 19.45731 (6.68) ***

PCNGRetailsites 47.84054 (3.14) ** PCTAFA 11.5088 (16.51) *** 4.9769 (27.91) *** 5.6332 (18.42) ***

SR 19.86695 (7.01) *** 7.709536 (14.84) *** 9.693157 (5.74) *** PCGDPPPP 0.0757(10.19) *** 0.0292 (14.03) *** 0.0415 (11.29) ***

PD 0.669348 (0.69) -0.43375 (1.38) 0.132692 (0.63) Years 10 (1999-2008) 10 (1999-2008) 10 (1999-2008)

No. of observations 480 480 480 Estimation method Parks Parks Parks

The numbers in the parentheses are the t-values.

+Significant at 0.1 level, *Significant at 0.05 level, ** Significant at 0.01 level, ***Significant at 0.001 level

26  

Table 7: TSCS analysis (N = 48)

Model 1 (Dependent variable: PCRetail)

Model 2 (Dependent variable: PCGRetail)

Model 3 (Dependent variable: PCNGRetail)

Intercept -640.899 (2.32) * -376.37 (3.84) *** -497.133 (5.02) *** ECFR 6.659 (1.75) + 5.56892 (3.59) *** 5.445688 (7.12) *** TSSC 6.999362 (2.67) ** 3.862723 (2.27) * 0.840776 (1.37)

PCFDIIS 904.4848 (0.23) 3567.754 (1.78) + 1759.067 (1.29) PCRetailsites -2.39529 (0.09)

PCGRetailsites -3.66469 (0.27) PCNGRetailsites 56.77669 (4.74) ***

PCTAFA 12.6324 (9.28) *** 5.3375 (15.01) *** 5.5983 (28.95) *** SR 32.22755 (3.06) ** 9.873048 (3.79) *** 15.00149 (4.81) ***

PCGDPPPP 0.074 (3.60) *** 0.0349 (8.84) *** 0.038153 (6.93) *** PD -0.12496 (0.28) 0.039457 (0.10) 3.274E - 7 (0.86)

Years 8 (1999-2006) 8 (1999-2006) 8 (1999-2006) No. of observations 384 384 384 Estimation method Parks Parks Parks

The numbers in the parentheses are the t-values.

+Significant at 0.1 level, *Significant at 0.05 level, ** Significant at 0.01 level, ***Significant at 0.001 level

27  

References Alexander, N. (1997), International Retailing. Oxford, UK: Blackwell Business.

Alexander, N. and Myers, H. (2000) “The retail internationalisation process”, International Marketing Review, 17(4/5), 334

Altrock, U. (1998) “The Lost Centre: Magnitogorsk Revisited”, Journal of Urban Design 3(2), 201-224.

Anderson, W., Chatterjee, L. and Lakshmanan, T. (2003) “E-commerce, Transportation, and Economic Geography”, Growth and Change, 34(4): 415-432.

Bass, F. and Wittink, D. (1975) “Pooling Issues and Methods in Regression Analysis with Examples in Marketing Research”, Journal of Marketing Research, 12 (November): 414-25.

Bater, J. (1980). The Soviet City (London, Edward Arnold).

Beauchamp, Michelle Bednarz; Ponder, Nicole. 2010. Perceptions Of Retail Convenience For In-Store And Online Shoppers, Marketing Management Journal, Vol. 20 Issue 1, p49-65

Blisard, Noel and Hayden Stewart (2007). Food Spending in American Households, 2003-04, Economic Information Bulletin No. (EIB-23) 115 pp, March 2007, http://www.ers.usda.gov/Publications/EIB23/

BMI (Business Monitor International) Romanians Continue Retail Spending Spree, November 2006, http://www.foodanddrinkinsight.com/file/40961/romanians-continue-retail-spending-spree.html

Bohata, M. (2000) "Skoda Automobilova a.s." in Saul Estrin, Xavier Richet and Joseph C. Brada, eds. Foreign Direct Investment in Central Eastern Europe: Case Studies of Firms in Transition. Armonk, New York: M. E. Sharp, pp. 158-175

Borchert, J. (1979) “Retail planning policy in the Netherlands”, in: R. L. Davies (Ed.) Retail Planning in the European Community, pp. 81-98. Farnborough: Saxon House.

Borchert, J. (1995) “Retail planning in the Netherlands”, in: R. L. Davies (Ed.) Retail Planning Policies in Western Europe, pp. 160-181. London: Routledge.

Branston, A. (2009). The Gulf Co-Operation Council Markets: A Retail Hotspot in a Cooling Economic Climate. Retail Digest, 44-49

Bronnenberg, B., Rossi, P. and Vilcassim, N. (2005) ”Structural Modeling and Policy Simulation”. Journal of Marketing Research, 42 (February): 22–26.

Buckley, D. R. (2007) “Economic integration a new approach to reform: The EBC Report on the Japanese Business Environment”, The European Business Council in Japan: 20-21.

Businessmonitor.,com (2007) The Latvia Food & Drink Report: The Sector At A Glance: Key Insights On The Food & Drink Sector Of Latvia, http://www.businessmonitor.com/food/latvia.html

Business Wire. 2005. Seven Multinational Companies Dominate the Global Retail Market for Hot Drinks, Which Was Worth... December 22 2005, http://www.allbusiness.com/company-activities-management/product-management/5185168-1.html

Chan, K. and McNeal, J. U. (2006) “Children and media in China: an urban-rural comparison study”, The Journal of Consumer Marketing 23(2), p. 79

Chandler, C. (2005) “The Great Wal-Mart of China”, Fortune, July 25, 104-110.

Chandler, C. (2007)“India's Global Reach”, Fortune, October 29, 156(9), 151-158

28  

Christaller, W. (1933): Die zentralen Orte in Süddeutschland. Gustav Fischer, Jena

Christian Science Monitor (1990) “Help Gorbachev through Joint Ventures”, August 21, 18

Christian Science Monitor (1999) “Not in the Same 'Big Box'”, Oct 6, 8

Chu, H. and Magnier, M. (2006). “The World; A Thaw That May Remake the World; After years of mistrust, growing powers India and China are gingerly developing relations”, Los Angeles Times. August 25, p. A.1.

Chu, J., Chintagunta, P. and Vilcassim, N. (2007) ”Assessing the Economic Value of Distribution Channels: An Application to the Personal Computer Industry”, Journal of Marketing Research, 44(1): 29-41.

Clarke, L, and Rimmer, P. (1997). “The anatomy of retail internationalisation: Diamaru's decision to invest in Melbourne”, Australia. Services Industries Journal 17:361-82.

Coe, N. M (2004) “The internationalization/globalization of retailing: Towards an economic geographical research agenda”, Environment and Planning 36:1571-94.

Coe, N. M. and Lee, Y. (2006) “The Strategic Localization of Transnational Retailers: The Case of Samsung-Tesco in South Korea”, Economic Geography, 82(1), 61-88.

Colla, E. and Dupuis, M. (2002) “Research and managerial issues on global retail competition: Carrefour/Wal-Mart”, International Journal of Retail & Distribution Management, 30(2), 103-111

Conger, A.J. 1974. A revised definition for suppressor variables: A guide to their identification and interpretation, Educational and Psychological Measurement, 34, 35-46.

Congressional Food Stamp Challenge, 2007. U.S. Members of Congress Live on a Food Stamp Budget May 21,. Speech on the Food Stamp Challenge, http://foodstampchallenge.typepad.com/my_weblog/congressman_jim_mcgovern/

Consumer Eastern Europe (1992), "Eastern Europe, a Market for the 1990s", Consumer Eastern Europe

Coulter, R., Price L. and Feick, L. (2003) “Rethinking the Origins of Involvement And Brand Commitment: Insights From Postsocialist Central Europe”, Journal of Consumer Research, 30(2): 151-169.

Country Commerce. 2006. The operating environment, China, 2006, 5-24.

Da Silva, J.G.C. (1975),"The Analysis of Cross-Sectional Time Series Data," Ph.D. dissertation, Department of Statistics, North Carolina State University.

Davis, D.S. and Sensenbrenner, J.S. (2000), "Commercializing childhood: parental purchases for Shanghai's only child", in Davis, D.S. (Ed.), The Consumer Revolution in Urban China, University of California Press, Berkeley, CA, pp. 54-79

Dickson, Peter R. 2000. Understanding the trade winds: The global evolution of production, consumption, and the Internet, Journal of Consumer Research. 27(1), 115-122.

Diehl, J. (1988) “Poland Eases Curbs on Private Sector; New Laws Are Seen as Major Step in Diluting Socialist System”, The Washington Post, December 24, p. a.01

Durand, C. (2007) “Externalities from foreign direct investment in the Mexican retailing sector”, Cambridge Journal of Economics 31(3), 393-411.

Dutt, E. (2005) “A.T. Kearney Study; India tops the 2005 Global Retail Development Index”, News India - Times. December 23, 36(51), p. 15.

Farrell, D. (2004) “The Case for Globalization”, The International Economy, 18(1): 52-55.

29  

Fomby, T., Hill, R. and Johnson, S. (1984). Advanced Econometric Methods, Springer-Verlag, New York.

Franses, P. (2005) “On the Use of Econometric Models for Policy Simulation in Marketing”, Journal of Marketing Research, 42 (February): 4–14.

French, R.A. and Hamilton, F.E. (Eds) (1979) The Socialist City (Chichester/New York/Brisbane/ Toronto, Wiley).

Fuller, W.A. and Battese, G.E. (1974), "Estimation of Linear Models with Crossed-Error Structure," Journal of Econometrics, 2, 67-78.

Gagnon, Marc-André (2007) “Capital, Power and Knowledge According to Thorstein Veblen: Reinterpreting the Knowledge-Based Economy,” Journal of Economic Issues, 41(2), 593-600.

Ganesh, J. (1998) “Converging Trends within the European Union: Insights from an Analysis of Diffusion Patterns”, Journal of International Marketing, 6(4): 32-48.

Ganesh, J., Kumar, V. and Subramaniam, V. (1997) “Learning Effect In Multinational Diffusion Of Consumer Durables: An Exploratory Investigation”, Academy of Marketing Science Journal, 25(3): 214-228.

Goldman, M. I. (2006) “Russia's Middle Class Muddle”, Current History, 105(693), 321-326.

Gumbel , P. (2006) “Comrades in Consumption”, Time International, May 15, p. 20.

Guy, C. M. (1998) “Controlling new retail spaces: The impress of planning policies in Western Europe”, Urban Studies. 35(5/6), 953-979.

Hanson, J. R II. (2003), “Proxies in the new political economy: Caveat emptor”, Economic Inquiry, 41(4), 639

Harper, R. (2002), The measurement of Social Capital in the United Kingdom, Office for National Statistics Paris, Organisation for Economic Co-operation and Development, September, http://www.oecd.org/dataoecd/22/52/2382339.pdf

Harris, Ethan S., Vega, Clara, 1996. What do chain store sales tell us about consumer spending?, Economic Policy Review, Vol. 2, Issue 2, 15-35.

Haslach, R. D. (1992). “The Bright Lights of Eastern Europe”, Europe, 314, 8-9.

Hilton, M. L. (2004) “Retailing The Revolution: The State Department Store (Gum) And Soviet Society In The 1920S”, Journal of Social History, 37(4), 939-965.

Hu, Z. L. (1993) Studies on Beijing's Central Business District, Urban Planning [Cheng Shi Gui Hua],l, pp. 45-47 (in Chinese).

Johnson, Jay L., 1995. Why American retailers are going global. Discount Merchandiser, Vol. 35, Issue 9, 37-38.

Kazmin, A. (2010). Indian market holds multinationals at bay, January 20, 2010 http://www.ft.com/cms/s/0/9d2bd1fe-05e9-11df-8c97-00144feabdc0.html#axzz1BG8ZQbKy

Keren, M. and Ofer, G. (2001) "FDI and The Restructuring of Trade and Supply Networks". Paper presented at the annual meeting of ACES, New Orleans, January,2001.

Keren, M. and Ofer, G. (2001) "The role of FDI in trade and financial services in transition: What distinguishes transition economies from developing economies?”, Comparative Economic Studies, 44(1), 15-45.

30  

Kilbinger, S. S. (2007) “Poland, Czech Republic See Retail Spurt as Spending Rises”, The Wall Street Journal Online, February 11, http://theestatenews.com/housing-market/Poland-Czech-Republic-See-Retail-Spurt-as-Spending-Rises/

King, W. B. (2007) “China Lowers the Bar for Foreigners”, Convenience Store News, 43(13), 101-104.

Kotabe, M. 2002. Using Euromonitor Database in International Marketing Research”, Academy of Marketing Science Journal, 30(2): 172-175.

Kotabe, M. and Helsen, K. (2001), Global Marketing Management, 2nd ed., John Wiley, New York.

Kshetri, N., Williamson, N. and Schiopu, A . (2007) “Economics and Politics of Advertising: Evidence from the Enlarging European Union”, European Journal of Marketing, 41(3/4): 349-66.

Kulke, E. (1992) “Structural change and spatial response in the retail sector in Germany”, Urban Studies, 29, pp. 965-977.

Kulke, E. (1996) “German fachmarkts-their effects on the structural and locational change in German retailing”, European Retail Digest, Spring, pp. 8-12.

Kulke, E. (1997) “Effects of the economic transformation process on the structure and locations of retailing in East Germany”, Journal of Retailing and Consumer Services, 4, pp. 49-55

Kumar, N. (1997) "The Revolution in Retailing: From Market Driven to Market Driving," Long Range Planning, 30 (6), 830-35.

Landes, D. S. ( 1969), The Unbound Prometheus: Technological Change and Industrial Development in Western Europe from 1750 to the Present, New York: Cambridge University Press.

Lau, K -N and Lam, P. (2002) “Economic freedom ranking of 161 countries in year 2000: a minimum disagreement approach”, The Journal of the Operational Research Society. 53(6), 664.

Lipman, M. (2007) “Russian Retail Politics”, The Washington Post. February 5, A.15

Los Angeles Times (2007) In Brief: Retail; Wal-Mart considering expansion into Russia, February 13, p. C.7

Maddala, G.S. (1971) “The Use Of Variance Components Models In Pooling Cross Section And Time Series Data”, Econometrica, 39(2), 341-58.

Mallaby, S. (1998) “Chalmers Johnson, Martin Feldstein, Francis Fukuyama. 1998. In Asia's mirror / Responses to Mallaby”, The National Interest, 52, 13-21.

Margolis, M. (2005) “It's A Mall World After All; Critics love to hate shopping centers as symbols of all that is wrong with American culture. But as they go global, they are stimulating economic growth throughout the developing world”, Newsweek, December 5, p. 28.

Mason, R.O. (1988) “Experimentation and Knowledge: A Pragmatic Perspective”, Knowledge, 10 (1): 3-24.

Morse, J. (1992). Qualitative Health Research, Newbury Park, CA, Sage Publications.

Næss, Petter. (2006). Accessibility, activity participation and location of activities: Exploring the links between residential location and travel behaviour Urban Studies, Mar2006, Vol. 43 Issue 3, p627-652

31  

Nikolayeva, Y. (2006) “Foreigners Will Be Forbbidden To Work As Salespeople”, The Current Digest of the Post - Soviet Press, December 13, 58(46), 15.

Pellet, Jennifer. 1995. Retailing in Eastern Europe, Discount Merchandiser, Vol. 35 Issue 8, 72-75.

Rai, S. (2006) “Retail Chains Scramble to Enter Indian Market”, New York Times, August 2, C.4

Reardon, T., Henson, S., and Berdegué, J. (2007) “Proactive fast-tracking' diffusion of supermarkets in developing countries: implications for market institutions and trade”, Journal of Economic Geography 7(4), 399-432.

Regmi, A. and Gehlhar, M. (2005) “Processed Food Trade Pressured by Evolving Global Supply Chains”, Amber Waves, 3(1), 12-19.

Reid, M. (1995) “Wet markets and warehouse clubs”, The Economist, March 4, 334(7904), SS14-15.

Robinson, S. (2007) “Food Fight”, Time International, June 11, p. 39.

Rose, Andrew K. 2004. Do WTO members have more liberal trade policy?, Journal of International Economics, 63(2), 209

Samiee, S. 1995. Strategic Considerations in European Retailing, Journal of International Marketing, 1995, Vol. 3 Issue 3, p49-76,

Samiee, S., Yip, L. S C. and Luk, S. T K (2004) “Viewpoint: International marketing in Southeast Asia: Retailing trends and opportunities in China”, International Marketing Review, 21(3), 247

Seiders, Kathleen,, Leonard L. Berry and Larry G. Gresham (2000), "Attention, Retailers! How Convenient Is Your Convenience Strategy?" Sloan Management Review, 41 (Spring), 79- 89.

Shy, Oz; Stenbacka, Rune. 2008. Price Competition, Business Hours and Shopping Time Flexibility, Economic Journal, Vol. 118 Issue 531, p1171-1195,

Spar, Karen. 2006. Cash and noncash benefits for persons with limited income: Eligibility rules, recipient and expenditure data, FY2002–2004. Report No. RL33340. Washington, DC: Congressional Research Service

Sternquist, Brenda (1997), "International Expansion of U.S. Retailers," International Journal of Retail and Distribution Management, 25 (8), 262-68.

The Economist (1991) “Asia's New Consumers: Pamper Them”, February 16, 56-57.

The Economist (1995) “ A new scramble”, August 12, 17-19.

The Economist (1999) “ Finance and economics: Surprise, surprise”, October 23, 353(8142), 88-89.

Trenkamp, Brad; Wiseman, Michael. 2007, The Food Stamp Program and Supplemental Security Income, Social Security Bulletin, Vol. 67 Issue 4, p71-87

Vega-Gordillo, M. and Alvarez-Arce, J. L (2003) “Economic Growth and Freedom: A Causality Study”, Cato Journal, 23(2), 199-215.

Vida, I and Fairhurst, A. (1998), "International expansion of retail firms: a theoretical approach for future investigations", Journal of Retail and Consumer Services, 5(3), 143-51.

32  

Vida, Irena; Fairhurst, Ann; Reardon, James. , 2000. Determinants of International Retail Involvement: The Case of Large U.S. Retail Chains . Journal of International Marketing, Vol. 8 Issue 4, p37-60

Walker, B. J., and Etzel, M. J. (1973) "The Internationalization of U.S. Franchise Systems: Progress and Procedures," Journal of Marketing 37, 38-46.

Wang, F. and Zhou, Y. (1999) “Modelling urban population densities in Beijing 1982-90: Suburbanisation and its causes”, Urban Studies, 36(2), 271-287.

Wrigley, N.. (2000a) “The globalization of retail capital: Themes for economic geography”, In The Oxford handbook of economic geography, ed. G. L. Clark, M. P. Feldman, and M. S. Gertler, 292-313. Oxford, U.K.: Oxford University Press.

Wrigley, N.. (2000b) “Strategic market behaviour in the internationalization of food retailing: Interpreting the third wave of Sainsbuiy's U.S. diversification”, European journal of Marketing 34:891-981.

Wu, J. (1993) “The historical development of Chinese urban morphology” Planning Perspectives, 8, pp. 20-52.

WWD: Women's Wear Daily, 2010. The Power Of Potential: China Lures U.S. Stores As Middle Class Grows. 6/24/2010, Vol. 199, Issue 132.

Yale, Laura and A. Venkatesh (1986), "Toward the Construct of Convenience in Consumer Research," Advances in Consumer Research, 13,403-408.

Yan, X. (1995) “Chinese urban geography since the late 1970s”, Urban Geography, 16, pp. 469-492.

33  

34  

                                                           

Notes:

 1 Analysts say that American retailers are not better by “birthright" but have faced a more competitive retail environment at home compared to their European and Japanese counterparts in their respective countries (Johnson, 1995). 2 In 1991, Belgium’s Delhaize Le Lion teamed up with the Czech Parik Group to open its first food supermarket in Prague. Similarly, Austrian food retailer, Julius Meinl, formed subsidiaries in Czech, Slovak, Poland, Hungary, Yugoslavia and Russia (Haslach, 1992). 3Because of the unavailability of FDI data specific to the retailing sector, we assumed that FDI in this sector is proportional to the total FDI in an economy. 4 The think-tank also moved to a 100 point system-- 100 indicating “perfectly free” (http://www.heritage.org/research/features/index/countries.cfm). 5 Even if random effects specification is found more appropriate on logical ground, one may still estimate fixed effects models. The fixed effect estimators are based on a particular sample which treats them as fixed in the sample (Hausman 1978).